9 Steps to Launch Your DevOps Initiative

,

You’ve recognized the benefits for your business with increased speed to market and cost savings by adopting a DevOps approach. And it sounds easy, doesn’t it?  “Let’s all work together.”  No problem. You’ve got this.

Start with the DevOps set of best practices.  What do you mean there isn’t one? Ok. Well, you can use the DevOps framework for adoption.  There isn’t one of those either?

Where DOES one start to  launch DevOps?

Step 1: Secure Executive Level Sponsorship

As in any organization transformation effort, visibility of your executive level sponsorship is critical, in-order to galvanize the buy-in required to affect the required cultural shift. As the DevOps sponsor you serve as the flag bearer that champions DevOps adoption within the organization, helps in negotiating financial commitment (budget), and serves as the final point of escalation.

If you are not in the proper position to accomplish these things, start your entire effort by identifying the executive that will appreciate the outlined benefits and serve in this critical role.

Step 2: Develop a Strategic Roadmap

Implementing a DevOps program requires the development of a high-level strategic roadmap in order to keep the organization on track through this continual improvement journey. Some of the important activities on the roadmap should include stakeholder-visioning workshops to understand key drivers for business agility, planning meetings to secure funding for automation and other critical adoption accelerators, and identifying key tactical milestones required to maintain stakeholder engagement.

Step 3: Understand “DevOps-Supported Services”

It is important to note that not all services in your portfolio will lend themselves to DevOps practices. As part of the DevOps program kick-off, conduct a comprehensive service portfolio review and business impact analysis to identify the services within your service portfolio that lend themselves to DevOps practices.

Rapidly evolving services hosted on modern technology platforms are prime candidates for early DevOps adoption, while legacy applications (e.g., mainframe-based applications) could experience more complex adoption challenges, as they could require platform migration.

As part of the due diligence, identify higher risk services to ensure risk mitigation strategies are in place for operational stability during the transformation.

Step 4: Adopt an Agile Software Development Methodology

Adoption of an Agile methodology for software development is a key requirement in-order to achieve the quick business responsiveness promoted by DevOps. This fast-looping and incremental software development approach consists of autonomous Scrum teams that work on product enhancements in sprint intervals. This allows for the continuous deployment of new features more quickly than a Waterfall methodology.

Agile development environments do not focus heavily on detailed requirement gathering and in-depth documentation for traceability, so a documentation strategy (e.g. leveraging the SDLC tool to capture detailed stories within a sprint or using RFC templates within the change management toolset) might be required in regulated entities for auditability purposes.

Step 5: Identify Supporting IT Processes to Integrate With the SDLC Lifecycle

In order to break down the wall and encourage a collaborative culture, you must identify key operational processes that should tightly integrate with the software development lifecycle and customize, as required, to make them fit for purpose.

Practical examples of processes to consider include:

Service Portfolio Management
This process should provide visibility into in-flight services in the pre-production pipeline.

Change & Release Management
These areas need to be lean and scalable enough to account for higher-frequency of deployment (e.g. peer review approvals).

Service Asset & Configuration Management (SACM)
This process should help identify the business impact and prioritize product deployment backlog.

Other examples include service operations processes such as incident, problem, knowledge, and event management. These areas should integrate early into the software development lifecycle in order to guarantee the appropriate level of support. This ensures service stability in a rapidly changing production environment. This will anchor your ability to rapidly identify root cause and to remediate problems within a minimal number of post-deployment sprints.

 

Step 6: Establish an Improvement Culture

A fast-looping continual service improvement culture is at the core of DevOps.

The technical teams (development and operations) must continually engage each other and the business stakeholders to:

  • Review and prioritize improvement opportunities
  • Conduct frequent post-deployment reviews to ensure that lessons learned are quickly fed back
  • Automate trend identification
  • Conduct frequent training workshops to educate stakeholders on
  • DevOps best practices
  • Audit/assess essential processes/functions at least once a year to identify opportunities for elimination of duplicate efforts and to drive continual service improvement

As part of the CSI culture, the organization must encourage innovation within the technical teams and promote the mind set for “fail faster”  to learn faster. This culture will foster speed and innovation. Technical teams will own failures and attack them as opportunities for improvement, versus sanctionable offences.

Step 7: Investment in Automation

Investing in a complementary suite of service automation technology is a major accelerator for DevOps adoption.

By leveraging innovative cloud-based solutions, organizations can automate complex, labor-intensive tasks and accelerate software deployment speed. By limiting the need for manual firefighting, you eliminate a huge source of potential error.

Use automated tools in the following areas:

  • Testing
  • Deployment
  • Service Monitoring
  • Application Performance Management
  • Instrumentation
  • Database/System/Server Monitoring
  • Network Monitoring

With a plethora of service automation solutions flooding the market, ensure that you are spending time assessing reliable solutions by referring to analyst reports authored by companies such as Forrester, Gartner, IDC or HsF.

Step 8: Establish a Communication Plan

At the heart of all DevOps is a well-orchestrated plan for communication of the effort and the expected benefits. Strategic communication will help break down the wall and encourage free flowing information. Good internal marketing also builds common vision, celebrates milestones and provides transparency on opportunities for improvement.

While developers and engineers operate in highly technical worlds, it is important to engage a communication expert to design less technical communications to demonstrate business values and keep non-technical stakeholders engaged.

Step 9: Develop Measurable Performance Metrics

As the phrase goes, “what gets measures gets improved.” Without a set of established performance baselines, it is impossible to demonstrate quantitative business values of DevOps.

Examples of DevOps (KPIs) key performance indicators include:

  • Stability (e.g. % of Incidents caused by Change, % of failed changes)
  • Scalability (e.g. % increase in weekly software release, % decrease in deployment backlog)
  • Reliability (e.g. MTBF, % increase in weekly defect volume)
  • Agility (e.g. % decrease in lead time for change deployment)
  • Responsiveness (e.g. MTTR)
  • Throughput (Volume of new changes deployed)

Ensure that whatever set of performance metrics you select, that each metric is measurable and can be distributed automatically.

 

SaveSave

0 replies

Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply

Your email address will not be published. Required fields are marked *

*